UCSPI

From: Jonathan de Boyne Pollard <J.deBoynePollard-newsgroups_at_NTLWorld.com>
Date: Mon, 08 Jun 2015 00:38:55 +0100

James Powell:
> For some comic relief:
> "We are systemd. You will be assimilated. Your projects and software
> will be deprecated, reinvented, and added into our own, and made to
> service us. Resistance is futile." -Lenncutus of Borg.

I don't really agree with all of the nonsense that surrounds this. But,
from the system release notes for version 220:

> When systemd forks off a new per-connection service instance
> it will now set the $REMOTE_ADDR environment variable to the
> remote IP address, and $REMOTE_PORT environment variable to
> the remote IP port. This behaviour is similar to the
> corresponding environment variables defined by CGI.

Those who don't learn about UCSPI are doomed to reinvent it badly, as here.

On that note, I am happy to report my discovery that GNU inetd has
UCSPI-TCP support.

* http://homepage.ntlworld.com./jonathan.deboynepollard/FGA/UCSPI.html

And I am interested to report my discovery that Laurent Bercot and I are
on one side of a design choice with Felix von Leitner and Erwin Hoffman
on the other.

And I am regretful and slightly hesitant to report a security bug in
s6-networking, where it fails (unless I have missed something) to wipe
any existing UCSPI-TCP environment variables that it isn't setting, per
the spec, but merely _merges_ environment changes.
Received on Sun Jun 07 2015 - 23:38:55 UTC

This archive was generated by hypermail 2.3.0 : Sun May 09 2021 - 19:44:19 UTC